Kinect

I was working on a framework to identify user(s) hand(s) gestures based on only depth image so that it can be deployed easily for anyone who owns a depth based camera(eg. Kinect, Asus Xtion)

Check out my prelimary video(Best viewed in HD):

I expect the use of this framework to be pretty extensive. Some examples I could think of are:

• Device-less presentation
• Power point presentation without going to your keyboard, giving you room to maneuver!
• Added functionality with being able to use gestures for more control. (eg. drawing, or some 3d visualisation)
• Interactive store display
• This is very useful as current interactive displays requires shops to place their touch screen display in the shop, or specially install the touch screen monitor at the shop front due to security reasons. If we replace it with this framework, we could just develop a software for the shopfront, and use any television and a Kinect pointing outside the shop. No special installation required. In addition, this set up can even catch curious passer-by when the set up reacts to them going past the shop.
• 3D Visualisation
• Use natural controls to navigate yourself in a 3D world instead of using a mouse and keyboard. Controlling a 3D environment with 3D controls by using the Kinect is way more intuitive than traditional keyboard and mouse.

I am currently in the works of creating a public version of Winect. If you are interested in helping out with the testing or any feedback, please let me know!